HP Cluster Test Software manual Nvidia Linpack Cuda Accelerated Linpack Benchmark

Page 25

Nvidia Linpack: CUDA Accelerated Linpack Benchmark

NOTE: For Nvidia GPUs only.

This test is implemented to run separately on each node using all CPU cores and all GPUs with about 80% or less of memory. Two parameters must be specified: Mem (GB) and Mem (%). Mem (GB) is the total memory in Giga Bytes on each node and Mem (%) is the percentage of total memory used in the test. This test does not give the cluster wide performance but it exercises CPU, GPU, and memory at the same time and it provides another set of numbers used to compare the relative performance among the cluster nodes. The amount of memory required is set to auto by default. This runs each node at the specified percentage of the memory available. For example, using auto and 80% a 2GB node uses 80% of 2GB while a 16GB node uses 80% of 16GB. The amount of memory used on any management node is adjusted to allow for increased memory consumption on management nodes.

Running accelerator tests 25

Image 25
Contents HP Cluster Test Administration Guide January Contents Sample test output Useful files and directories Utility commandsDocumentation feedback Glossary Index Varieties of Cluster Test CT Image using a networkCT Image RPM Starting Cluster Test Cluster Test GUIFiles generated by Cluster Test Running cluster testsCluster Test GUI Running cluster tests Configuration settings Running tests in a batch Using scripts to run tests Running cluster tests CrissCross Test descriptionsNodes monitoring window Monitoring tests and viewing resultsTest output window Monitoring tests and viewing results Performance analysis Test report Checking the InfiniBand fabric Cluster Test toolbar menus Cluster Test toolbar menusStarting accelerator tests Accelerator test GUIFiles generated by accelerator test GPU detection Running accelerator testsVerify Dgemm Double Precision General Matrix Multiply Test Sgemm Single Precision General Matrix Multiply TestBandWidth GPU Bandwidth Test Memory TestNvidia Linpack Cuda Accelerated Linpack Benchmark Configuring Cluster Test when using RPM Cluster Test procedure as recommended by HPAdditional software Accelerator test procedure Cluster Test procedure as recommended by HP Cluster Test procedure # checkadm Cluster Test procedure Cluster Test procedure as recommended by HP Performance monitor utility Performance monitorPerformance Monitor toolbar menu Xperf utility Hardware Inventory Cluster Test toolsFirmware Summary Server health check Excluding the head node from tests Disk Scrubber Cluster Test tools Running tests in parallel An example per-node directory Creating and changing per node filesAn example cloned per-node directory Nfs NFS performance tuningDetecting new hardware TroubleshootingCluster Test Troubleshooting Guide Troubleshooting Cluster TestScope of this document Support and other resourcesIntended audience Contacting HPRelated information New and changed information in this editionDocumentation WebsitesCustomer self repair Typographic conventionsCustomer self repair Cluster Test Useful Files and Directories Useful files and directoriesAnalyze Utility commandsConrep Files generated by ibfabriccheck Inspectibfabric.pl Inspectibfabric.pl Utility commands Ipmitool Pdsh CrissCross Sample test outputSample test output Test4 Pallas Mpibyte Sample test output Stream Node24 Triad 3078.7949 3355 3488 3536 CPU Disk TestUTK LinpackPassed Passed Passed Documentation feedback CMU GlossaryIndex MPI Accelerator
Related manuals
Manual 25 pages 60.17 Kb